# OpenMobile-2 trains phone agents in simulated apps
> OpenMobile-2's authors report a gain on a separate physical-phone benchmark.
By BIGCHANGE.AI Editorial
Published: 2026-10-26T06:11:40.021Z
Updated: 2026-10-26T06:11:40.021Z
Canonical: https://bigchange.ai/blog/openmobile-2-simulated-app-training-real-phone-test

AI-generated conceptual illustration by BIG CHANGE.
An [October 9 preprint](https://arxiv.org/html/2610.14157v1) reports resettable app training and a separate physical-phone test.
## The big change
- **What changed:** The team built 35 commercial-style simulated apps with checkable state.
- **Why it matters:** Controlled tasks let researchers compare screen actions and experimental app-native tools; they do not prove live-app utility.
- **What to watch:** One physical-phone benchmark tests transfer, not everyday reliability.
## The measured boundary
MobileGym++ checks 215 simulated tasks. On SPA-Bench L3 physical phones, the 27B model scored 57.4% after hybrid supervised training and 59.6% after reinforcement learning: a 2.2 percentage-point gain. BIG CHANGE did not run either test.
## Sources
- [Paper v1](https://arxiv.org/html/2610.14157v1) — Author results.
- [Project](https://os-copilot.github.io/OpenMobile2-Home/) — Artifacts.
- [Code](https://github.com/OpenMobile2/OpenMobile2-Code) — Methods and licensing.
- [Simulator](https://github.com/OpenMobile2/MobileGym-plusplus) — Tasks and terms.
The BIG CHANGE newsletter
The big picture. At your pace.
Recent stories on AI and robotics, the shifts worth watching and practical ideas to use. Choose a daily briefing, weekly digest or monthly perspective.
Next scheduled send (UTC): . Your first edition arrives at the next scheduled send after you confirm.
Your privacy, your choice.
Necessary storage supports site security and remembers your choices. Optional Google Analytics stays off until you allow it. You can read every story with necessary storage only. Privacy details